Regional POI demand identification method based on multi-source feature fusion collaborative filtering

Abstract

The invention discloses a regional POI demand identification method based on multi-source feature fusion collaborative filtering. The method comprises the following steps: 1, acquiring regional related data and POI related data; 2, designing an MR access inference algorithm based on K neighbors to obtain regional trajectory data of POI accessed by a user; 3, analyzing and processing the region related data, the POI related data and the region track data of the POI accessed by the user into a form which can be input by a neural network; 4, constructing a nerve collaborative filtering model with an attention mechanism; 5, optimizing a nerve collaborative filtering model with an attention mechanism; 6, modeling a relationship between the region and the POI to obtain a POI demand of each region. According to the method, a multi-feature fusion collaborative filtering means is adopted, the crowd trajectory is considered, the geographic features of the region and the evaluation features of the POI are combined, the relationship between the region and the POI is modeled through the neural collaborative filtering model, the algorithm complexity is low, and the demand analysis precision is high.

Description

technical field [0001] The invention relates to the technical field of data services, in particular to a regional POI requirement identification method based on multi-source feature fusion and collaborative filtering. Background technique [0002] With the development of cities, the need to identify POIs (Point-Of-Interest) in urban areas is crucial to the construction of smart cities, such as town planning and commercial location selection. Regarding this problem, the traditional demand identification algorithm based on user survey reports is less robust and has a strong lag. The current mainstream regional POI demand identification is mainly designed for POIs with certain characteristics. Although this type of prediction is more practical, it has poor versatility. It can only model some specific POIs, and the model cannot expanded to other fields. Contents of the invention [0003] Aiming at the deficiencies of the above-mentioned prior art, the present invention provi...

AI Technical Summary

Problems solved by technology

Regarding this problem, traditional demand identification algorithms based on user survey reports are less robust and have a strong lag

At present, the mainstream regional POI demand identification is mainly designed for a certain type of POI. Although this type of prediction is more practical, it has poor versatility. It can only model some specific POIs, and the model cannot expand to other fields

Images